A contract theory approach to business bankruptcy.

• Author: Schwartz, Alan

Business bankruptcy systems attempt to solve a coordination problem for the creditors of insolvent firms. Some insolvent firms cannot earn revenues sufficient to cover nonfinancing costs. The assets of these firms can be put to better uses elsewhere, so the assets should be sold off.(1) Other insolvent firms earn revenues that exceed production costs but that are too low to service the firm's debt. These firms should be continued as going concerns under less leveraged capital structures. As is well known,(2) an individual creditor is not interested in this distinction: The creditor would rather maximize returns by attaching assets sufficient to pay its claim in full. If every creditor attempts to attach assets, the firm will be liquidated piecemeal, whether it is efficient to continue the firm or not.

Creditors as a group would prefer to coordinate their collection actions so that the debtor firm is liquidated piecemeal only when that would raise more money for creditors than continuing the firm--"reorganizing" it--would raise. Creditors are widely believed to have high coalition costs, however, and there is credible evidence for this view: Private workouts after insolvency often fail.(3) Western bankruptcy systems commonly respond to this creditors' problem by enacting coordination mechanisms. These mechanisms prevent some or all creditors from collecting debts individually and establish procedures to make the value-maximizing choice between liquidation and reorganization. The value of the insolvent firm is then distributed to creditors who participate in the bankruptcy procedure according to a priority scheme.

Western countries require the debtor and many (sometimes all) creditors to participate in the state-supplied bankruptcy system and restrict the ability of 4 parties to alter certain outcomes that the state system directs.(4) To appreciate these peculiar bankruptcy features, recall that the typical commercial dispute is resolved by a court but parties can contract for a different dispute resolution procedure, such as arbitration. The typical body of commercial law also is a set of defaults; the rules govern unless particular parties choose-different rules in their contract.(5) In contrast, parties cannot contract in lending agreements to use a bankruptcy system other than the one the state supplies.(6) Also, parties cannot modify many of the rules that constitute the system. For example, the Uniform Commercial Code (U.C.C.) permits a party to cancel a deal if its contract partner becomes insolvent,(7) and parties have commonly made the Code's authorization explicit by contracting for a right to exit.(8) Today, bankruptcy law permits the insolvent party (or its bankruptcy trustee) to keep a sales contract in force despite state law or private agreement.(9)

That bankruptcy systems solve a coordination problem rather than regulate the substance of transactions accounts for some of the distinctions between bankruptcy and commercial law generally. For example, if contracting were otherwise unregulated, a contractual term authorizing creditors to collect promptly and in full upon default should not be enforced because enforcement of such terms would result in piecemeal liquidations. Structural rules of the game must be mandatory or the game cannot be played at all. Bankruptcy systems, however, contain more mandatory rules than a structural account can explain.

This Essay makes three claims regarding the restrictive features of modern bankruptcy systems. The first claim follows from the recognition that the optimality of a bankruptcy system is state-dependent. Reorganizing an insolvent firm sometimes would maximize revenue, while liquidating the same firm would be best under other values of the relevant economic parameters. The United States recognizes the state-dependent optimality of bankruptcy systems by authorizing a court to decide ultimately whether an insolvent firm should be liquidated or reorganized.(10) This Essay shows that parties could improve on this solution with contracts that induce the use of the system that is optimal in their particular circumstances. These contracts, therefore, should be legal.

The Essay's second claim is that the only mandatory rules in a bankruptcy system should be structural. The current U.S. Bankruptcy Code contains mandatory structural rules as well as mandatory rules whose goal is to augment the value of the bankrupt estate. The rule barring the solvent party from canceling a contract with the insolvent party is such a rule: It requires the solvent party to perform at a possible loss if the deal is profitable to the estate.(11) There is, however, no good distributional reason to benefit one set of business creditors--the "estate"--at the expense of another set--the solvent party. Nonstructural mandatory bankruptcy rules are justifiable only if they increase ex post efficiency. The cancellation rule, therefore, should require the solvent party to perform only when the estate's gain from performance would exceed the solvent party's loss.

It does not follow, however, that bankruptcy law should attempt to increase ex post efficiency with mandatory rules. The Coase Theorem teaches that, when certain conditions hold, the initial location of a property right, as set by mandatory rules, is irrelevant to efficiency. Bankruptcy systems create mechanisms to facilitate Coasean bargaining. These systems often appoint a state official to represent the bankrupt estate,(12) and the official can bargain with creditors, such as the solvent party in the example here. Nonstructural mandatory bankruptcy rules are thus needed only when the conditions for Coasean bargaining do not obtain. This Essay shows that bargaining succeeds with respect to money: Renegotiation after insolvency will shift tangible wealth to the estate when it is efficient to do so but not otherwise, whether or not a mandatory rule is present. Because bankruptcy systems should function to maximize the monetary value of the estate,(13) these systems need not contain mandatory redistributional rules. Therefore, the only mandatory rules in a bankruptcy system should be structural.

The two claims just summarized implicitly assume that bankruptcy systems exist only to increase efficiency by solving the creditors' coordination problem. Many American commentators argue that bankruptcy systems also should protect persons or entities who do not have current claims against the insolvent firm. In the literature, protected classes include workers with an interest in continued employment and local communities that benefit from the firm's continued presence.(14) Perhaps more of the American system's mandatory character could be justified if the system were meant to protect nonparties, but this Essay's third claim is that the better arguments hold that bankruptcy systems should solve only the creditors' coordination problem.(15)

The importance of this Essay's topic is evidenced by recent commercial practice and scholarly comment. Business parties have become increasingly dissatisfied with the U.S. system and are attempting to avoid it despite apparent legal prohibitions. The National Bankruptcy Review Commission observed:

While it was long ago assumed that specific rights, effects, or obligations

provided by the Bankruptcy Code could not be waived in advance even in the

absence of an express nonwaivability Code provision, case law and business

practice have begun to call this long-held assumption in question. With

increasing regularity, loan documents and workout agreements contain

clauses waiving the automatic stay [of creditor collection rights] if

the borrower files for bankruptcy. (16)

The Commission added: "Other typical clauses provide that filing a [bankruptcy] petition will constitute 'bad faith' if intended to forestall foreclosure, or that the debtor agrees to admit the existence of facts that will support a case dismissal order."(17) The Commission responded to these efforts to avoid bankruptcy by recommending that the Bankruptcy Code be amended to make explicit that "a clause in a contract or lease ... does not waive, terminate, restrict, condition, or otherwise modify any rights or defenses provided by Title 11" of the Code.(18)

Scholars, meanwhile, have made three recent contributions to the debate about the mandatory character of bankruptcy systems. These contributions are consistent with this Essay's claim that parties should be free to choose preferred bankruptcy systems in their lending agreements. First, some scholars argue that requiring parties to use the state-supplied system is not a serious constraint. Firms may avoid any such system by selecting "bankruptcy-proof" capital structures.(19) For example, firms that eschew debt cannot go bankrupt. Firms, however, choose capital structures to solve agency problems between the firm's managers or owners and outside investors.(20) Firms thus face the difficult problem of choosing between a suboptimal capital structure that would avoid bankruptcy or reduce bankruptcy costs and an otherwise optimal capital structure that will compel the firm, if insolvent, to use a suboptimal bankruptcy system. This choice could be avoided if firms were free to contract directly for a preferred bankruptcy procedure.

Second, some scholars have argued that a firm should be permitted to choose its preferred bankruptcy system in its corporate charter. Creditors of the firm would be bound by the firm's charter choice.(21) This solution may founder over the problem that because the optimality of a bankruptcy system is state-dependent, it also can be time-dependent. Put more simply, time may render a company's charter solution outmoded, and corporate charters are inconvenient to amend. Some firms would likely prefer the relative flexibility of using lending agreements to induce the choice of optimal bankruptcy systems.

Finally, some courts and scholars argue that firms should...